Hit-to-Lead Optimization and Discovery of 5-((5-([1,1′-Biphenyl]-4-yl)-6-chloro-1H-benzo[d]imidazol-2-yl)oxy)-2-methylbenzoic Acid (MK-3903): A Novel Class of Benzimidazole-Based Activators of AMP-Activated Protein Kinase

Article abstract of DOI:10.1021/acs.jmedchem.7b01344

AMP-activated protein kinase (AMPK) plays an essential role as a cellular energy sensor and master regulator of metabolism in eukaryotes. Dysregulated lipid and carbohydrate metabolism resulting from insulin resistance leads to hyperglycemia, the hallmark of type 2 diabetes mellitus (T2DM). While pharmacological activation of AMPK is anticipated to improve these parameters, the discovery of selective, direct activators has proven challenging. We now describe a hit-to-lead effort resulting in the discovery of a potent and selective class of benzimidazole-based direct AMPK activators, exemplified by 5-((5-([1,1′-biphenyl]-4-yl)-6-chloro-1H-benzo[d]imidazol-2-yl)oxy)-2-methylbenzoic acid, 42 (MK-3903). Compound 42 exhibited robust target engagement in mouse liver following oral dosing, leading to improved lipid metabolism and insulin sensitization in mice.

Full text of DOI:10.1021/acs.jmedchem.7b01344

Subscriber access provided by Eastern Michigan University | Bruce T. Halle Library
Article
Hit-to-lead Optimization and Discovery of 5-((5-([1,1'-
Biphenyl]-4-yl)-6-chloro-1H-benzo[d]imidazol-2-yl)oxy)-2-
methylbenzoic Acid (MK-3903): A Novel Class of Benzimidazole-
based Activators of AMP-activated Protein Kinase
Ping Lan, F. Anthony Romero, Dariusz Wodka, Andrew J Kassick, Qun Dang, Tony Gibson,
Dan Cashion, Gaochao Zhou, Yuli Chen, Xiaoping Zhang, Aihua Zhang, Ying Li, Maria
E. Trujillo, Qing Shao, Margaret Wu, Shiyao Xu, Huaibing He, Deidre MacKenna, Jocelyn
Staunton, Kevin T Chapman, Ann Weber, Iyassu K Sebhat, and Gergely M. Makara
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01344 • Publication Date (Web): 16 Oct 2017
Downloaded from http://pubs.acs.org on October 16, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Hitꢀtoꢀlead Optimization and Discovery of 5ꢀ((5ꢀ
([1,1'ꢀBiphenyl]ꢀ4ꢀyl)ꢀ6ꢀchloroꢀ1Hꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
benzo[d]imidazolꢀ2ꢀyl)oxy)ꢀ2ꢀmethylbenzoic Acid
(MKꢀ3903): A Novel Class of Benzimidazoleꢀbased
Activators of AMPꢀactivated Protein Kinase
Ping Lan,*† F. Anthony Romero,† Dariusz Wodka,† Andrew J. Kassick,† Qun Dang,∆ Tony
Gibson,∆ Daniel Cashion,∆ Gaochao Zhou,‡ Yuli Chen,‡ Xiaoping Zhang,‡ Aihua Zhang,‡ Ying
Li,‡ Maria E. Trujillo,ǁ Qing Shao,ǁ Margaret Wu,ǁ Shiyao Xu,# Huaibing He,# Deidre
MacKenna,∆ Jocelyn Staunton,∆ Kevin T. Chapman,† Ann Weber,† Iyassu K. Sebhat*† and
Gergely M. Makara†
†Department of Medicinal Chemistry, ‡Department of Biology, ǁDepartment of Pharmacology,
#Department of Pharmacokinetics, Pharmacodynamics, and Drug Metabolism, MRL, Merck &
Co., Inc., 2000 Galloping Hill Road, Kenilworth, New Jersey 07033, United States.
∆Metabasis Therapeutics, Inc., 11119 North Torrey Pines Road, La Jolla, California 92037,
United States.
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 48
1
2
3
4
5
6
7
8
9
ABSTRACT: AMPꢀactivated protein kinase (AMPK) plays an essential role as a cellular energy
sensor and master regulator of metabolism in eukaryotes. Dysregulated lipid and carbohydrate
metabolism resulting from insulin resistance leads to hyperglycemia, the hallmark of type 2
diabetes mellitus (T2DM). While pharmacological activation of AMPK is anticipated to improve
these parameters, the discovery of selective, direct activators has proven challenging. We now
describe a hitꢀtoꢀlead effort resulting in the discovery of a potent and selective class of
benzimidazoleꢀbased direct AMPK activators, exemplified by 5ꢀ((5ꢀ([1,1'ꢀbiphenyl]ꢀ4ꢀyl)ꢀ6ꢀ
chloroꢀ1Hꢀbenzo[d]imidazolꢀ2ꢀyl)oxy)ꢀ2ꢀmethylbenzoic acid, 42 (MKꢀ3903). Compound 42
exhibited robust target engagement in mouse liver following oral dosing, leading to improved
lipid metabolism and insulin sensitization in mice.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
INTRODUCTION
Since its discovery in 1973,1,2 extensive research has contributed to our understanding of the
biology of 5'ꢀadenosine monophosphateꢀactivated protein kinase (AMPK) including its
fundamental role as a major sensor of cellular energetics and regulator of metabolism in
eukaryotes.3,4,5 AMPK is a conserved serine/threonine protein kinase. Activation of AMPK in
liver inhibits fatty acid synthesis (FAS) and stimulates fatty acid oxidation (FAO) by
phosphorylating and inactivating both isoforms of acetyl CoA carboxylase ACC1 and ACC2, the
first enzymes shown to be downstream targets for AMPK.6.7 ACCs are rateꢀlimiting enzymes for
the synthesis of malonylꢀCoA, which is a potent inhibitor of fatty acid oxidation and a precursor
in fatty acid biosynthesis.8 Activation of AMPK in primary culture of hepatocytes was shown to
reduce gene expression of the key gluconeogenic enzymes, phosphoenolpyruvate carboxykinase
(PEPCK) and glucoseꢀ6ꢀphosphatase (G6Pase).9,10 In skeletal muscle, AMPK activation
ACS Paragon Plus Environment
2

Page 3 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
stimulates glucose uptake acutely because of increased translocation of the glucose transporter
(GLUT4) to the plasma memberane.11 Gene expression of GLUT 4 is upꢀregulated via AMPK
activation as well.12 Dysregulation of metabolism in man leads to metabolic syndrome, a
condition characterized by insulin resistance. Elevated ectopic intracellular lipids are considered
a major cause of insulin resistance. If left untreated, insulin resistance leads to hyperglycemia,
which defines type 2 diabetes mellitus (T2DM), due in part to the inability of insulin to suppress
hepatic glucose production.13 The combined metabolic changes in lipid and glucose homeostasis
induced by AMPK activation are anticipated to contribute to increased metabolic flexibility and
insulin sensitivity. Indeed, insulin sensitivity of muscle, liver and whole body was increased after
a single dose of 5ꢀaminoimidazoleꢀ4ꢀcarboxamideꢀ1ꢀβꢀDꢀribofuranoside (AICAR), a widely
used phosphorylated AMP anologue for AMPK activation, in insulinꢀresistant rat model.14
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Conversely, there are potential concerns associated with systemic AMPK activation, in
particular, with AMPK activation in the heart .15,16 Mutations in the AMPK γ2ꢀsubunit lead to
PRKAG2 cardiomyopathy. These mutations appear to prevent the binding of inhibitory ATP to
the allosteric nucleotideꢀbinding regulatory site, resulting in nonꢀphysiological activation of
AMPK.17 Given the potential association of adverse effects with systemic AMPK activation,
AMPK isoformꢀselective or tissueꢀselective activation appeared an attractive approach for
developing activators.
A number of agents, including two antihyperglycemic drugs, have been reported to activate
AMPK in cells, tissues and/or in vivo.10, 18 While the detailed modes of action for many of these
agents may not be fully delineated, most of them activate AMPK in an indirect manner.19,20 The
discovery of direct activators of AMPK has proven to be challenging, in part due to its structural
complexity. The enzyme is a heterotrimer comprised of a catalytic (α1 or α2), a “scaffold” (β1 or
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 48
1
2
3
4
5
6
7
8
9
β2) and a regulatory (γ1 or γ2 or γ3) subunit, all of which are encoded by separate genes.
Excluding splice variants, there are a total of 12 AMPK complexes in mammals.21 It is not clear
which isoform(s) are required to realize the maximum benefits of AMPK activation in vivo.
Furthermore, distinct complex/subunit distributions exist in different tissues of human and other
preclinical animal species. For example, human skeletal muscle contains predominately α2β2ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
containing complexes.22,23 The β1ꢀcontaining complexes, including α1β1γ1 and α2β1γ1, account
for 95% of total AMPK activity in rodent liver extract.24 However recently published reports
suggest α1β2γ1 to be the major isoform in human liver.25,26 Despite major efforts to identify
direct AMPK activators, few molecules have been reported.27,28,29 Some of the more widely used
compounds include thienopyridone 1 (Aꢀ769662),30 the first smallꢀmolecule direct AMPK
activator with β1ꢀselectivity, the phosphonic acid 2 (identified from a library of AMP
mimetics)31 and PT1ꢀrelated AMPK activator 3.32
Figure 1. Representative AMPK direct activators
Repeated attempts to identify direct AMPK activators through traditional highꢀthroughputꢀ
screens (HTS) of the MSD compound collection failed to generate progressable hits. The team
therefore opted to take a less traditional approach involving the bioassay of a proprietary
fragment collection (informer library) at high concentration. Using this approach, we identified
several hits, established the tractability of the SAR of a hit bearing a benzimidazole core (Figure
ACS Paragon Plus Environment
4

Page 5 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1, compound 4), efficiently optimized that series, and ultimately discovered suitable tool
compounds for proofꢀofꢀconcept studies, all in less than one year. These efforts, as well as the
results of initial biological evaluation of an optimized tool compound 42 (MKꢀ3903)33 are now
detailed here.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
RESULTS AND DISCUSSION
Following multiple unsuccessful HTS campaigns to identify direct AMPK activators, we
conducted an informer library screen to find nontraditional chemotypes as a starting point for
optimization efforts. This informer library comprised a subset of our internal compound
collection, with molecular weights of >200 Da and containing up to 22 heavy atoms. The heavy
atom count was used as the molecular weight upper filter to avoid biasing against heavier
elements such as chlorine. The in silico construction of the library utilized typical
physicochemical property filters such as solubility and logP. The collection was intentionally
designed to straddle the boundary between fragmentꢀlike and leadꢀlike libraries.34 By biasing the
deck towards larger and less hydrophilic fragments than those typically used in NMRꢀ or Xꢀrayꢀ
based screens, we hoped to increase the likelihood of identifying hits with <250 ꢀM potency
using a bioassay performed with high compound concentrations. The collection consisted of
~25,000 small molecules. These were incorporated as singletons, without close analogs, in order
to maximize the coverage of chemical space and to represent a wide variety of scaffolds.
For the bioassay, AMPK activity was assessed by measuring the phosphorylation of the
synthetic peptide substrate SAMS by recombinant human AMPK complex 7 (α2β1γ1). The
α2β1γ1complex was preꢀactivated by phosphorylation of the catalytic α subunit to form
phosphorylated AMPK (pAMPK).21 Further activation of pAMPK by the allosteric activator
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 48
1
2
3
4
5
6
7
8
9
AMP served as the positive control and comparator for the efficacy of our hits. Three hits (>30%
activation vs. maximal activation attained by AMP) were identified from the informer screen.
One of these was the benzimidazole derivative 5,6ꢀdichlorobenzimidazoleꢀ2ꢀpropionic acid with
an EC50 of ~30 ꢀM (compound 4, Table 1).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The low molecular weight and micromolar potency of 4 translated to a ligand efficiency index
(LE)35 of 0.38, which we considered a good starting point for optimization. An investigation of
the SAR of hit 4 was therefore initiated, beginning with an examination of the effect of aromatic
ring substitution. Compounds 7 ꢀ 11 were synthesized following known procedures (Scheme
1).36 Commercially available phenylenediamines 5 with various substituents were condensed
with succinic anhydride at high temperature. The amide intermediates 6, when heated with acetic
acid, cyclized to give the benzimidazole 2ꢀpropionic acids. For compounds 12 and 13, the 4ꢀ
chloroꢀ5ꢀbromobenzimidazole (intermediate 11) was further derivatized via Suzuki crossꢀ
coupling37 with the corresponding boronic acids. All analogs were assayed against pAMPK 7
and the maximum activation at a concentration of 50 ꢀM is shown in Table 1. Compounds
without substitution (7) or with substituents such as methoxy (8) or small alkyl groups (9)
showed very little activation. The presence of a single chlorine atom on the aromatic ring (10)
was sufficient to allow some degree of activation, which underscored the importance of this
substituent. When this chlorine was retained and the other position was substituted with a phenyl
ring (12), a modest (~3ꢀfold) activity gain was realized vs. 4. An additional small potency
enhancement was observed upon replacement of the phenyl ring with a biphenyl group (13).
Scheme 1. Synthesis of 3-(1H-benzo[d]imidazol-2-yl)propanoic Acidsa
ACS Paragon Plus Environment
6

Page 7 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Effect of 3-(1H-benzo[d]imidazol-2-yl)propanoic Acids on pAMPK7 Activity
% maximum
EC50 (ꢀM)b
Compound
R1
Cl
R2
Cl
H
AMP
activationa
78%
14%
14%
19%
41%
44%
130%
130%
36
NDc
NDc
NDc
NDc
NDc
8
4
7
H
OMe
Me
Cl
H
8
Me
H
9
10
11
12
13
Cl
Br
Cl
Cl
6
a %max. activation referred to activation level at 50 ꢀM. bValues are the average of at least 2
experiments, each using a 10ꢀpoint titration. cNot Determined.
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 48
1
2
3
4
5
6
7
8
9
In parallel, some of our efforts were directed towards modifications in the benzimidazole 1ꢀ
and 2ꢀpositions (Scheme 2). A complete loss of activity was observed when the benzimidazole ꢀ
NH was methylated (data not shown). A number of benzimidazoles containing modified acid
linkages were quickly prepared via the condensation of oꢀphenylenediamine 14 and the
corresponding arylaldehydes (for 16, 17), or via nucleophilic substitution of bromoꢀsubstituted
alkylcarboxylic acids by the thiol moiety in 5,6ꢀdichloroꢀ1Hꢀbenzimidazoleꢀ2ꢀthiol 18 (for 19,
21) (Table 2). Compounds where benzoic acids were directly attached to the benzimidazole 2ꢀ
position (16, 17) lost activity. Interestingly, introduction of a sulfur atom in the 2ꢀposition linker
(compound 19) resulted in significant improvement in both maximum activation percentage and
potency relative to 4. Oxidation of the thio group to a sulfoxide (20) led to a loss of activity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Syntheses of 3-(1H-benzo[d]imidazol-2-yl) Acids with Various Linkersa
Table 2. Effect of (1H-benzo[d]imidazol-2-yl) Acids with Various Linkers on pAMPK7
Activity
% maximum
EC50 (ꢀM)b
Compound
AMP
activationa
ACS Paragon Plus Environment
8

Page 9 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
6%
NDc
NDc
16
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10%
17
CO2H
Cl
Cl
N
130%
8.5
S
N
H
19
CO2H
Cl
Cl
N
15%
NDc
S
O
N
H
20
130%
4
21
a %max. activation referred to activation level at 50 ꢀM. bValues are the average of at least 2
experiements, each using a 10ꢀpoint titration. cNot Determined.
We expanded the SAR by looking at additional acid linkers containing the newly discovered
thio substituent. The route used for previous analogs (Scheme 2) was unsuitable for this purpose
so we looked to develop a more efficient synthesis that would allow lateꢀstage incorporation of
the benzimidazole 2ꢀsubstituent. As illustrated in Scheme 3, benzimidazoleꢀ2ꢀthiol 18 was
converted to 2ꢀmethylsulfonylbenzimidazole 22 via methylation with methyl iodide followed by
oxidation with mꢀCPBA. The desired 2ꢀarylthio benzimidazoles were obtained following
reaction of 22 with various substituted arylthiophenols.38 Generally, arylthio substitution was
well tolerated. Acid substitution in the metaꢀ or orthoꢀpositions (26, 27) resulted in significantly
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 48
1
2
3
4
improved potency compared to the initial hit and homologation of the acid appeared to be well
5
6
tolerated (25 vs. 27).
7
8
9
Scheme 3. Synthesis of 1H-benzo[d]imidazol-2-yl)thioaryl Acidsa
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Cl
Cl
Cl
Cl
Cl
Cl
O Me
S
R
N
N
N
a, b
c
SH
S
N
H
N
H
N
H
O
23-27
18
22
aReagents and conditions: (a) MeI, K2CO3, acetone, rt; (b) m-CPBA, DCM, rt; (c) arylthiols,
EtOH, 70oC, followed by hydrolysis: LiOH, MeOH/THF/H2O, 40oC, when needed.
Table 3. Effect of 1H-benzo[d]imidazol-2-yl)thioaryl Acids on pAMPK7 Activity
% maximum
EC50 (ꢀM)b
Compound
23
R
AMP
activationa
35%
76%
NDc
2.6
2.9
24
25
140%
130%
150%
8.1
7.4
26
27
a %max. activation referred to activation level at 50 ꢀM. bValues are the average of at least 2
experiments, each using a 10ꢀpoint titration. cNot Determined.
ACS Paragon Plus Environment
10

Page 11 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
We next examined whether the potency gains realized from our investigation of SAR in the
benzimidazole 2ꢀ and 5ꢀpositions would be additive. The key intermediates 32 and 33 provided
ideal starting points to quickly access the fullyꢀassembled analogs (35 to 38) via standard crossꢀ
coupling conditions (Scheme 4). The resultant hybrid compounds had significantly improved
potency with EC50 values in the nanomolar range. The most potent analog, containing a 2ꢀ
hydroxylbiphenyl group (38) exhibited an EC50 of 29 nM. While less potent, compound 35
activated the complex 320% versus the %maximal activation observed with the physiological
activator AMP.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 4. Synthesis of (6-Chloro-5-aryl-1H-benzo[d]imidazol-2-yl)thioacetic Acids &
Thiobenzoic Acidsa
Table 4. Effect of (6-Chloro-5-aryl-1H-benzo[d]imidazol-2-yl)thioacetic Acids &
Thiobenzoic Acids on pAMPK7 Activity
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 48
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
% maximum AMP
activationb
EC50 (ꢀM)a
0.20
Compound
Ar
R
320%
260%
200%
100%
35
36
37
38
0.15
0.80
0.029
aValues are the average of at least 2 experiments, each using a 10ꢀpoint titration. b %max.
activation referred to the highest activation level observed in the titration concentration range (~2
nM – 50 ꢀM).
An important metabolic consequence of AMPK activation is the siteꢀspecific phosphorylation/
inhibition of ACC. This leads to the inhibition of fatty acid synthesis, also known as de novo
lipogenesis (DNL).39 Measures of compound effects on DNL thus served as a useful, proximal
pharmacodynamic marker of AMPK target engagement and activation. Administration of 30
mg/kg compound 36 to mice intraperitoneally (i.p.) led to ~40% inhibition of DNL. A much
higher oral dose of 100 mg/kg oral was required for more modest reductions in DNL (~20%).
Pharmacokinetic (PK) studies established that compound 36 had high clearance (60 mL/min/kg)
and poor bioavailability (7%) (Table 5). Examination of Pꢀglycoprotein 1 (Pꢀgp) susceptibility
and permeability of the compound demonstrated very high transport in Lilly laboratories porcine
kidney (LLCꢀPK1) control cells. These data implicated 36 as a substrate of another transporter
which was subsequently identified as the breast cancer resistance protein (BCRP/ABCG2).
ACS Paragon Plus Environment
12

Page 13 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
BCRP is expressed in the luminal plasma membrane of enterocytes, where it is responsible for
the efflux of substrates into the intestine. The transporter is also expressed in hepatocytes where
it pumps substrates into bile.40 When compound 36 was dosed in BCRP knockꢀout (KO) mice, a
~30ꢀfold reduction in clearance and commensurate ~10ꢀfold increase in bioavailability was
observed, suggesting that BCRP plays an important role in the elimination and/or absorption of
36.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 5. PK Parameters for Compound 36 in Wild-type and Bcrp Knock-out Micea
WT
60
KO
2.9
CLp (mL/min/kg)
t1/2 (h)
2.8
7
10.8
77
F (%)
Vdss (L/kg)
6.7
2.6
AUCN (PO) (uM*h*kg/mg)
0.05
11.4
aCLp: plasma clearance; t1/2: terminal halfꢀlife; F: bioavailability; Vdss: volume of distribution at
steady state; AUCN (PO): normalized area under plasma concentration vs time curve following
PO dosing.
Given the identified transporter liabilities, we continued with efforts to optimize structures. To
this end, we examined the utility of methylsulfone intermediate 34 to access 2ꢀaryloxy analogs.
Initial attempts to directly apply this approach were unsuccessful, but modified solventꢀfree
conditions utilizing elevated reaction temperatures afforded the desired products in modest yields
(Scheme 5, step b) .38 An alternate route was developed later using SEMꢀprotected
benzimidazole 39, which afforded higher yields under milder conditions. Using both approaches,
we quickly synthesized a group of Oꢀlinked analogs with various aryl or alkyne groups in the
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 48
1
2
3
4
5
6
benzimidazole 5ꢀposition. This effort delivered potent AMPK activators with single and doubleꢀ
digit nanomolar EC50 values.
7
8
9
Scheme 5. Synthesis of 5,6-Substituted ((1H-benzo[d]imidazol-2-yl)oxy)benzoic Acidsa
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
I
R
R
Me
O
Me
O
Ar
N
N
N
b
a
S
S
O
N
H
N
H
N
H
O
O
Cl
Cl
Cl
34
33
41-48
Me
e, f
CO2Me
c
I
I
N
Me
O
N
N
d
O
S
N
Cl
O
Cl
SEM
SEM
40
39
aReagents and conditions: (a) corresponding arylboronic acid or ester, Pd(PPh3)4, K3PO4, dioxane,
100oC, or corresponding alkynes, CuI, PdCl2(PPh3)2, Et3N, DMF, 120oC; (b) corresponding
methyl 3-hydroxybenzoate, neat, 130oC, followed by hydrolysis; (c) Et3N, SEM-Cl,
THF, rt; (d) methyl 5-hydroxy-2-methylbenzoate, K2CO3, DMF, rt; (e) corresponding arylboronic ester,
Pd(PPh3)4, K2CO3, DMF, 120oC; (f) TBAF, THF, 80oC, followed by 2.5 N aq. NaOH, 45oC.
Table 6 details the data for the resulting compounds and highlights the impact of single atom
modifications. The replacement of the sulfur for an oxygen atom resulted in at least one order of
magnitude gain in activity (e.g., ~13ꢀfold gain for compound 41 vs. 37). Further enhancement of
activity was accomplished when small substituents were added to the benzoic acid moiety. For
example, the introduction of a paraꢀCl or paraꢀMe led to ~2ꢀ7ꢀfold improvement in potency (42
and 43 vs. 41). In line with previous SAR, the introduction of the hydroxy group at the ortho
position of the biphenyl group furnished compounds with further improved potency (47).
Smaller benzimidazole 5ꢀsubstituents were also tolerated, with the fused heterocyclic
methylindole group affording one of the most potent activators (48). A crystal structure of
compound 48, in complex with human pAMPK, was recently published.41 This study helped to
ACS Paragon Plus Environment
14

Page 15 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
define a novel allosteric activation site on pAMPK and provided detailed structural information
that helps to explain the potency of the series.
7
8
9
Table 6. Effect of 5,6-Substituted ((1H-benzo[d]imidazol-2-yl)oxy)benzoic Acids on
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
pAMPK7 Activity
% maximum AMP
EC50 (nM)a
Compound
R
Ar
activationb
61
9
260%
41
320%
310%
240%
250%
260%
230%
340%
42
43
44
45
46
47
48
32
7
15
11
3
2
aValues are the average of at least 2 experiments, each using a 10ꢀpoint titration. b %max.
activation referred to the highest activation level observed in the titration concentration range
(~0.8 nM – 16 ꢀM).
As part of our optimization efforts, we monitored ligand efficiency metrics. While the high LE
of the hit compound was well conserved, the progression of lipophilic ligand efficiencies (LLEs)
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 48
1
2
3
4
5
6
7
8
9
remained relatively unchanged or slightly decreased for later compounds. LLE is an index
combining both in vitro potency and lipophilicity.35 The concurrent increase in potency and
lipophilicity suggested that the binding affinity improvement realized with later compounds
came mainly from enhanced lipophilic interactions between the compounds and the putative
binding sites.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Ligand efficiencies during hitꢀtoꢀlead progression
Compound 42 was selected, together with other candidate compounds, for further in vitro and
in vivo profiling. Activation was assessed for all 12 recombinant human pAMPK complexes.
Compound 42 activates 10 of the 12 pAMPK complexes with EC50 values in the range of 8ꢀ40
nM and maximal activation >50% (Table 7.). The compound partially activates pAMPK5 (36%
Max), which is only a minor component of human and mouse liver, and it does not activate
pAMPK6 which is not detected in liver.24
ACS Paragon Plus Environment
16

Page 17 of 48
Journal of Medicinal Chemistry
1
2
3
4
Table 7. Effect of Compound 42 at the 12 Recombinant Human pAMPK Complexes
5
6
7
8
Human pAMPK
EC50 (nM)(%max)a
Subunit composition
complex
9
1
2
3
4
5
6
7
8
9
8 (217%)
21 (242%)
19 (61%)
40 (50%)
118 (36%)
2860 (8%)
9 (320%)
22 (175%)
26 (209%)
39 (94%)
23 (145%)
35 (135%)
α1 β1 γ1
α1 β1 γ2
α1 β1 γ3
α1 β2 γ1
α1 β2 γ2
α1 β2 γ3
α2 β1 γ1
α2 β1 γ2
α2 β1 γ3
α2 β2 γ1
α2 β2 γ2
α2 β2 γ3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10
11
12
aEC50 values are the average of at least 2 experiments, each using a 10ꢀpoint titration. %max.
referred to the highest activation level observed in the titration concentration range (~2 nM – 50
ꢀM).
The pharmacokinetics of 42 in C57BL/6 mice, SpragueꢀDawley rats and beagle dogs were
characterized by moderate systemic plasma clearance (5.0 ꢀ 13 mL/min/kg), a volume of
distribution at steady state of 0.6 ꢀ1.1 L/kg and a terminal halfꢀlife of ~2 hr (Table 8). Compound
42 had low oral bioavailability (8.4%) in C57BL/6 mice, but the oral exposure was later
improved using other vehicles (data not shown). Oral bioavailability in rats and dogs was
improved (27 ꢀ 78%).
Table 8. Pharmacokinetic Parameters of Compound 42 in Pre-clinical Speciesa
Pharmacokinetic Parameter
SPECIES
Rat
Mouse
2/10
6.8
Dog
0.5/1
13
Dose IV/PO (mg/kg)
CLp (mL/min/kg)
1/4
5.0
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 48
1
2
3
4
5
6
7
8
9
Vdss (L/kg)
t1/2 (hr)
0.7
2.1
2.0
1.0
8.4
0.5
0.6
1.7
4.8
0.5
25
1.1
2.4
1.1
0.5
78
C
max (ꢀM)
tmax (hr)
F (%)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2.6
2.3
Oral AUCN (ꢀM•hr•kg/mg)
aCLp: plasma clearance; Vdss: volume distribution at steady state; t1/2: terminal halfꢀlife; Cmax
maximum concentration; tmax: time of maximum concentration; F: bioavailability; AUCN:
normalized area under plasma concentration vs time curve.
:
Compound 42 demonstrated low permeability (Papp = 6 x10ꢀ6 cm/s) in LLCꢀPK1 cells,42 and
is a substrate of human liver uptake transporters OATP1B1 and OATP1B3 (organic anion
transporter proteins).43 42 is a weak reversible inhibitor of CYP3A4 and 2D6 in human liver
microsomes (apparent IC50 >50 ꢀM) and did not exhibit timeꢀdependent inhibition of CYP3A4
activity. In the in vitro hPXR assay 42 had an EC50 of >30 ꢀM suggesting that it is not a potent
PXR agonist.44 The Panlab screen and follow up evaluation showed that 42 binds moderately to
the prostanoid DP2 (CRTH2) receptor (binding IC50 = 1.8 ꢀM) but not in the presence of 10%
human serum (binding IC50 > 86 ꢀM).45 42 was also screened at Upstate against a panel of
kinase targets, identifying p38ꢀregulated/activated protein kinase (PRAK) as the only hit with
IC50 < 10 ꢀM.
As discussed above, the level of phosphorylated ACC (pACC), a downstream AMPK
substrate, is an indicator of intracellular AMPK activation. The pACC to total ACC ratio,
measured with a novel Meso Scale Discovery (MSD) assay, was used as a proximal target
engagement readout for AMPK activation in vivo. Diet induced obese (DIO) mice (n=8) were
treated with 42 (30 mg/kg BID) for 15 days. Liver and muscle were harvested 2 hours following
ACS Paragon Plus Environment
18

Page 19 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
the final dose. pACC/ACC ratios were significantly increased in both liver (3.1ꢀfold) and skeletal
muscle (1.6ꢀfold) (Figure 3).
7
8
9
a. DIO Liver
b. DIO Muscle
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Effect of compound 42 (30 mg/kg BID, 15 days) administration on pACC/ACC in (a)
liver and (b) muscle of DIO mice 2 h after the final dose
In addition, we measured inhibition of de novo fatty acid synthesis (FAS) in liver, one
functional consequence of increased ACC phosphorylation. Acute oral administration of
compound 42 (3, 10, and 30 mg/kg) to highꢀfructose fed db/+ mice resulted in significant
inhibition of hepatic FAS for all three doses. Specifically, db/+ mice were fed for 1 week on a
diet enriched with 64% fructose to elevate rates of FAS. FAS was measured both 2 hours and 8
hours after compound dosing. In this dose titration study, 42 inhibited hepatic FAS by 60%,
61%, and 66% 2 hours after doses of 3 mg/kg, 10 mg/kg, and 30 mg/kg. At 8 hours after dosing,
inhibition of FAS by 42 was 29%, 37%, and 49% at these respective doses (Figure 4.).
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 48
1
2
3
4
5
6
7
8
9
250
200
150
100
50
2 hours
8 hours
**
***
60%
***
61%
*
***
66%
29%
***
37%
49%
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0
Figure 4. Acute effects of 42 on hepatic fatty acid synthesis in db/+ mice. *p < 0.05 compared to
vehicle; **p < 0.01 compared to vehicle; ***p < 0.001 compared to vehicle.
Compound 42 was also assessed for its chronic effects on insulin sensitization in DIO mice.
Figure 5. Chronic effect of compound 42 on a measure of insulin sensitivity in DIO mice
following 12 days of dosing. oGTT IRI = measured glucose (mM) x measured insulin (ng/ml).
*p < 0.05; **p < 0.01; ***p < 0.001.
ACS Paragon Plus Environment
20

Page 21 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
DIO Mice are normoglycemic but hyperinsulinemic, serving as a good model of insulin
resistance or prediabetes.46 DIO mice treated with compound 42 (3ꢀ30 mg/kg BID and 30 mg/kg
QD p.o.) for 12 days were subjected to an oral glucose tolerance test (oGTT) following a 4 hour
fast on day 12. The effects on blood glucose and insulin 20 min after bolus glucose
administration were combined to provide an insulin resistance index (IRI) (Figure 5.).47 A dose
dependent reduction in IRI was observed, mainly driven by suppression of insulin excursion
during the oGTT. There were no significant effects on blood glucose. A modest effect on body
weight at 30 mg/kg BID (3.1% loss of body weight) complicated interpretation of the data for
that dose group, but significant effects on IRI were observed without body weight change for the
10 mg/kg BID and 30 mg/kg QD groups. Taken together, these data suggest that compound 42
holds promise as an insulin sensitizing agent.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CONCLUSION
A biochemical informer (fragment) library screen was used to identify a novel class of AMPK
activators where prior HTS approaches had failed. Starting from a low affinity and low
molecular weight hit, we were able to optimize the benzimidazole pharmacophore by
successfully utilizing both replacement and extension strategies. The first breakthrough came
from the replacement of the 2ꢀbenzimidazole carbon linkage with thiobenzoic acid, which
effectively propelled the potency of the resulting analogs into the low micromolar range. The
second significant improvement was realized when the benzimidazole core was extended at the
5ꢀposition, appending a hydrophobic moiety that nearly doubled the size of the structures. The
gain in size was crucial for manifesting potency, facilitating interaction with a highly
hydrophobic pocket generated at the interface between the carbohydrateꢀbinding module of the
βꢀsubunit and the kinase domain of the αꢀsubunit.41 The last potency boost came from a single
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 48
1
2
3
4
5
6
7
8
9
atom replacement of sulfur with oxygen at the benzimidazole 2ꢀposition, furnishing compounds
with single digit nanomolar potencies. Overall, these optimization efforts led to four orders of
magnitude improvement in potency versus the original hit and ~4ꢀfold increase in the %maximal
activation achieved by the physiological activator AMP. These efforts identified molecules
suitable for in vivo biological studies. Chronic oral administration of compound 42 robustly
increased ACC phosphorylation in liver with more modest effects in skeletal muscle. Treatment
of various mouse models with compound 42 resulted in expected alterations in lipid metabolism
and improvements in a measure of insulin sensitization. Based in part on these promising
preclinical results, compound 42 was designated as a development candidate to support further
research on this novel class of AMPK activators. The results of the extensive follow up studies
employing 42 and other compounds derived from this class of AMPK activators will be reported
separately.48
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EXPERIMENTAL SECTION
General Experimental Methods: Commercially available starting materials, reagents, and
solvents were as used as received. Anhydrous reactions were performed under an inert
atmosphere such as nitrogen. Microwave reactions were performed using a Biotage Initiator
microwave reactor. Reaction progress was generally monitored by LCMS (Waters Acquity
UPLC). Flash column chromatography purification of intermediates was performed on an ISCO
CombiFlashRf system using prepacked ISCO silica gel columns. Elution was performed using a
gradient of hexane/ethyl acetate. Preparative reverseꢀphase HPLC purification of final products
was performed on a Waters AutoPurification System with a Waters CSH (Cꢀ18, 100 mm x 19
mm, 5ꢀm) reverseꢀphase column. Elution was performed using a gradient of acetonitrile/water
(0.16% TFA) over an 8ꢀmin period at a flow rate of 25 mL/min. The fractions were detected with
ACS Paragon Plus Environment
22

Page 23 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
PDA and MS detectors. Fractions containing the desired material were concentrated using
GeneVac HTꢀ24 to obtain the final products. The purity analyses of all final compounds were
performed on Waters Acquity UPLC equipped with BEHꢀC18 column (50 mm × 1.0 mm, 1.7
ꢁm), using gradient 10:90 to 95:5 v/v CH3CN/H2O + v 0.05% TFA over 1.6ꢀmin period at a flow
rate of 0.3 mL/min, with UV wavelength 254 nm. All tested compounds exhibited >95% purity
under the LC conditions. Proton NMRs were recorded on a Bruker 500 MHz spectrometer, and
were analyzed using MestReNova program. Chemical shifts are expressed in ppm and referenced
to deuterated dimethyl sulfoxide (DMSOꢀd6) or acetone (acetoneꢀd6) or methanol (CD3OD). All
reported yields are not optimized.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
General procedure for preparation of 3-(1H-benzo[d]imidazol-2-yl)propanoic acids. To a
solution of benzeneꢀ1,2ꢀdiamine (0.2 mmol, 1 equiv) in NMP (1 mL) was added succinic
anhydride (0.2 mmol, 1 equiv). The mixture was stirred for overnight at 80oC. Then acetic acid
(1 mL) was added, and the afforded solution was stirred for overnight under reflux. The reaction
was concentrated to dryness, and the desired product was isolated by preparative HPLC.
3-(5,6-Dichloro-1H-benzo[d]imidazol-2-yl)propanoic acid (4). The title compound was
prepared in the manner similar to the general procedure by utilizing 4,5ꢀdichlorobenzeneꢀ1,2ꢀ
diamine as the starting material. 1H NMR (500 MHz, Acetoneꢀd6) δ 7.69 (s, 2H), 3.17 (t, J = 7.0
Hz, 2H), 2.90 (t, J = 7.0 Hz, 2H) .
3-(1H-Benzo[d]imidazol-2-yl)propanoic acid (7). The title compound was prepared in the
manner similar to the general procedure by utilizing benzeneꢀ1,2ꢀdiamine as the starting
material. 1H NMR (500 MHz, Acetoneꢀd6) δ 7.49 (dd, J = 6.0, 3.2 Hz, 2H), 7.14 (dd, J = 6.0, 3.2
Hz, 2H), 3.17 (t, J = 7.0 Hz, 2H), 2.90 (t, J = 7.0 Hz, 2H).
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 48
1
2
3
4
5
6
7
8
9
3-(5-Methoxy-1H-benzo[d]imidazol-2-yl)propanoic acid (8). The title compound was
prepared in the manner similar to the general procedure by utilizing 4ꢀmethoxybenzeneꢀ1,2ꢀ
diamine as the starting material. 1H NMR (500 MHz, Acetoneꢀd6) δ 7.38 (d, J = 8.7 Hz, 1H),
7.03 (d, J = 2.4 Hz, 1H), 6.78 (dd, J = 8.7, 2.4 Hz, 1H), 3.79 (s, 3H), 3.14 (t, J = 7.0 Hz, 2H),
2.87 (t, J = 7.0 Hz, 2H).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3-(5,6-Dimethyl-1H-benzo[d]imidazol-2-yl)propanoic acid (9). The title compound was
prepared in the manner similar to the general procedure by utilizing 4,5ꢀdimethylbenzeneꢀ1,2ꢀ
diamine as the starting material. 1H NMR (500 MHz, CD3OD) δ 7.31 (s, 2H), 3.16 (d, J = 7.3
Hz, 2H), 2.81 (d, J = 7.8 Hz, 2H), 2.35 (s, 6H).
3-(5-Chloro-1H-benzo[d]imidazol-2-yl)propanoic acid (10). The title compound was
prepared in the manner similar to the general procedure by utilizing 4ꢀchlorobenzeneꢀ1,2ꢀ
diamine as the starting material. 1H NMR (500 MHz, DMSOꢀd6) δ 7.86 (dd, J = 2.0, 0.6 Hz, 1H),
7.77 (dd, J = 8.7, 0.6 Hz, 1H), 7.53 (dd, J = 8.7, 2.0 Hz, 1H), 3.30 (t, J = 7.1 Hz, 2H), 3.01 (t, J =
7.1 Hz, 2H).
3-(6-Bromo-5-chloro-1H-benzo[d]imidazol-2-yl)propanoic acid (11). The title compound
was prepared in the manner similar to the general procedure by utilizing 4ꢀbromoꢀ5ꢀ
chlorobenzeneꢀ1,2ꢀdiamine as the starting material (10 mg, 8% yield). 1H NMR (500 MHz,
DMSOꢀd6) δ 7.85 (s, 1H), 7.73 (s, 1H), 3.00 (t, J = 7.1 Hz, 2H), 2.68 (t, J = 7.1 Hz, 2H).
3-(5-Chloro-6-phenyl-1H-benzo[d]imidazol-2-yl)propanoic acid (12). 3ꢀ(6ꢀbromoꢀ5ꢀchloroꢀ
1Hꢀbenzo[d]imidazolꢀ2ꢀyl)propanoic acid (0.39 mmol, 1 equiv) (11) and phenylboronic acid (1
equiv) were dissolved in anhydrous dioxane (3 mL), then catalyst, Pd(PPh3)4 (5 mol%), was
added followed by 1M K2CO3 (2 equiv). The resulting mixture was stirred at 80oC overnight.
ACS Paragon Plus Environment
24

Page 25 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The reaction was concentrated to dryness, and the desired product was isolated by preparative
HPLC (38 mg, 32% yield). 1H NMR (500 MHz, DMSOꢀd6) δ 7.87 (s, 1H), 7.62 (s, 1H), 7.48ꢀ
7.39 (m, 5H), 3.20 (t, J = 7.1 Hz, 2H), 2.88 (t, J = 7.1 Hz, 2H).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3-(6-([1,1'-Biphenyl]-4-yl)-5-chloro-1H-benzo[d]imidazol-2-yl)propanoic acid (13). The
title compound was prepared in the manner similar to compound 12 by utilizing [1,1'ꢀbiphenyl]ꢀ
4ꢀylboronic acid as the starting material (69 mg, 47% yield). 1H NMR (500 MHz, DMSOꢀd6) δ
7.91 (s, 1H), 7.79ꢀ7.77 (m, 2H), 7.75ꢀ7.73 (m, 2H), 7.70 (s, 1H), 7.56ꢀ7.54 (m, 2H), 7.51ꢀ7.48
(m, 2H), 7.40ꢀ7.37 (m, 1H), 3.22 (t, J = 7.1 Hz, 2H), 2.90 (t, J = 7.1 Hz, 2H).
Preparation of 2-(5,6-dichloro-1H-benzo[d]imidazol-2-yl)benzoic acid (16). To a solution of
4,5ꢀdichlorobenzeneꢀ1,2ꢀdiamine (0.2 mmol, 1 equiv) in ethanol (2 mL) and water (1 mL) was
added methyl 2ꢀformylbenzoate (1 equiv) followed by sodium metabissulfite (2 equiv). The
mixture was stirred at 80oC for overnight, and then was concentrated to dryness. LiOH (10
equiv) was added to the afforded residue, followed by addition of solvents MeOH, THF and
water (1 mL each). The mixture was stirred at 40oC. After overnight, the reaction was
concentrated to dryness, and the desired product was isolated by preparative HPLC (10 mg, 16%
yield). 1H NMR (500 MHz, DMSOꢀd6) δ 8.76 (t, J = 1.7 Hz, 1H), 8.41ꢀ8.39 (m, 1H), 8.07 (dt, J
= 7.8, 1.4 Hz, 1H), 7.87 (s, 2H), 7.70 (t, J = 7.8 Hz, 1H).
3-(5,6-Dichloro-1H-benzo[d]imidazol-2-yl)benzoic acid (17). The title compound was
prepared in the manner similar to the procedure as 16 by utilizing methyl 3ꢀformylbenzoate as
the starting material (7 mg, 12% yield). 1H NMR (500 MHz, DMSOꢀd6) δ 8.76 (t, J = 1.7 Hz,
1H), 8.40 (dt, J = 7.8, 1.5 Hz, 1H), 8.07 (dt, J = 7.8, 1.4 Hz, 1H), 7.87 (s, 2H), 7.69 (t, J = 7.8
Hz, 1H).
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 48
1
2
3
4
5
6
7
8
9
Preparation of 3-((5,6-dichloro-1H-benzo[d]imidazol-2-yl)thio)propanoic acid (19). To a
solution of 5,6ꢀdichloroꢀ1Hꢀbenzo[d]imidazoleꢀ2ꢀthiol (18) (0.2 mmol, 1 equiv) in ethanol (1
mL) was added 3ꢀbromopropanoic acid (1 equiv) and KOH (2 equiv). The mixture was stirred
overnight at rt. The reaction was concentrated to dryness, and the desired product was isolated by
preparative HPLC (14 mg, 20% yield). 1H NMR (500 MHz, DMSOꢀd6) δ 7.70 (s, 2H), 3.42 (t, J
= 6.9 Hz, 2H), 2.76 (t, J = 6.9 Hz, 2H).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3-((5,6-Dichloro-1H-benzo[d]imidazol-2-yl)sulfinyl)propanoic acid (20). Compound 3ꢀ
((5,6ꢀdichloroꢀ1Hꢀbenzo[d]imidazolꢀ2ꢀyl)thio)propanoic acid (19) (0.2 mmol, 1 equiv) was
dissolved in DCM (4 mL), then mꢀCPBA (1 equiv) was added. The mixture was stirred at rt
overnight. The reaction was concentrated to dryness, and the desired product was isolated by
preparative HPLC (20 mg, 30% yield). 1H NMR (500 MHz, Acetoneꢀd6) δ 7.90 (s, 2H), 3.70ꢀ
3.64 (m, 1H), 3.43 (ddd, J = 13.7, 7.9, 6.4 Hz, 1H), 2.93ꢀ2.87 (m, 1H), 2.66 (ddd, J = 17.5, 7.9,
6.4 Hz, 1H).
2-((5,6-Dichloro-1H-benzo[d]imidazol-2-yl)thio)acetic acid (21). The title compound was
prepared in the manner similar to the procedure as 19 by utilizing 2ꢀbromoacetic acid as the
starting material (32 mg, 50% yield). 1H NMR (500 MHz, DMSOꢀd6) δ 7.69 (s, 2H), 4.15 (s,
2H).
Preparation of intermediate 5,6-dichloro-2-(methylsulfonyl)-1H-benzo[d]imidazole (22).
Step1: 5,6ꢀdichloroꢀ2ꢀ(methylthio)ꢀ1Hꢀbenzo[d]imidazole. K2CO3 (14 mmol, 2 equiv),
followed by iodomethane (8.4 mmol, 1.2 equiv), was added to a solution of 18 (7 mmol, 1 equiv)
in acetone (20 mL) at 0oC. The reaction was stirred at rt overnight. Volatiles were removed and
the residue was partitioned between EtOAc and water. Concentration afforded the desired
ACS Paragon Plus Environment
26

Page 27 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
product as a white foam, which was subjected to chromatography on a column of SiO2 eluting
with a gradient from 30% to 100% EtOAc in hexanes, to provide the title compound.
7
8
9
Step 2: 5,6ꢀdichloroꢀ2ꢀ(methylsulfonyl)ꢀ1Hꢀbenzo[d]imidazole. mꢀChloroperbenzoic acid (10
mmol, 2 equiv) was added to a suspension of 5,6ꢀdichloroꢀ2ꢀ(methylthio)ꢀ1Hꢀbenzo[d]imidazole
(5 mmol, 1 equiv) in DCM (50 mL). The reaction was stirred at rt for overnight then washed
with 1N aqueous NaOH. The organic phase was concentrated. The residue was subjected to
chromatography on a column of SiO2 eluting with a gradient from 10% to 100% EtOAc in
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
hexanes, to provide the title compound (500 mg, 41% yield). H NMR (500 MHz, DMSOꢀd6) δ
8.00 (s, 2H), 3.51 (s, 3H).
General procedure for preparation of 1H-benzo[d]imidazol-2-yl)thioaryl acids. To a
solution of intermediate 22 (0.2 mmol, 1 equiv) in ethanol (1.5 mL) was added the corresponding
mercaptobenzoic acid (1 equiv). The mixture was stirred at 70oC overnight. The reaction was
concentrated to dryness, and the desired product was isolated by preparative HPLC.
4-((5,6-Dichloro-1H-benzo[d]imidazol-2-yl)thio)benzoic acid (23). The title compound was
prepared in the manner similar to the general procedure by utilizing 4ꢀmercaptobenzoic acid as
the starting material (26 mg, 40% yield). 1H NMR (500 MHz, DMSOꢀd6) δ 7.95ꢀ7.93 (m, 2H),
7.79 (s, 2H), 7.59ꢀ7.57 (m, 2H).
2-((5,6-Dichloro-1H-benzo[d]imidazol-2-yl)thio)thiazole-4-carboxylic acid (24). The title
compound was prepared in the manner similar to the general procedure by utilizing ethyl 2ꢀ
mercaptothiazoleꢀ4ꢀcarboxylate as the starting material, followed by hydrolysis with addition of
a solution of LiOH (5 equiv) in mixed solvents MeOH/THF/water for 2 h (14 mg, 20% yield). 1H
NMR (500 MHz, CD3OD) δ 8.45 (s, 1H), 7.75 (s, 2H).
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 48
1
2
3
4
5
6
7
8
9
2-(3-((5,6-Dichloro-1H-benzo[d]imidazol-2-yl)thio)phenyl)acetic acid (25). The title
compound was prepared in the manner similar to the general procedure by utilizing 2ꢀ(3ꢀ
mercaptophenyl)acetic acid as the starting material (34 mg, 48% yield). 1H NMR (500 MHz,
Acetoneꢀd6) δ 7.63ꢀ7.61 (m, 3H), 7.54ꢀ7.51 (m, 1H), 7.43ꢀ7.40 (m, 2H), 3.68 (s, 2H).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2-((5,6-Dichloro-1H-benzo[d]imidazol-2-yl)thio)benzoic acid (26). The title compound was
prepared in the manner similar to the general procedure by utilizing 2ꢀmercaptobenzoic acid as
the starting material (32 mg, 47% yield). 1H NMR (500 MHz, DMSOꢀd6) δ 7.96 (dd, J = 7.8, 1.6
Hz, 1H), 7.86 (br s, 1H), 7.43 (ddd, J = 8.1, 7.3, 1.6 Hz, 1H), 7.33 (td, J = 7.5, 1.2 Hz, 1H), 6.92
(dd, J = 8.1, 1.1 Hz, 1H).
3-((5,6-Dichloro-1H-benzo[d]imidazol-2-yl)thio)benzoic acid (27). The title compound was
prepared in the manner similar to the general procedure by utilizing 3ꢀmercaptobenzoic acid as
the starting material (27 mg, 41% yield). 1H NMR (500 MHz, DMSOꢀd6) δ 8.04 (t, J = 1.8 Hz,
1H), 7.98ꢀ7.96 (m, 1H), 7.81 (ddd, J = 7.8, 2.0, 1.1 Hz, 1H), 7.75 (s, 2H), 7.59ꢀ7.56 (m, 1H).
Preparation of intermediate tert-butyl 2-((6-chloro-5-iodo-1H-benzo[d]imidazol-2-
yl)thio)acetate (32).
Step 1: 5ꢀChloroꢀ4ꢀiodoꢀ2ꢀnitroaniline (29). To a solution of 5ꢀchloroꢀ2ꢀnitroaniline 28 (145
mmol, 1 equiv) in AcOH (250 mL) was added Nꢀiodosuccinimide (145 mmol, 1 equiv). The
mixture was stirred overnight at 50oC, cooled down to rt and filtered. The solid residue was
washed with AcOH, water, saturated aqueous NaHCO3 and water, and then dried to afford the
desired product as a brown solid, which was used in the next step without further purification.
ACS Paragon Plus Environment
28

Page 29 of 48
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Step 2: 4ꢀChloroꢀ5ꢀiodobenzeneꢀ1,2ꢀdiamine (30). To a suspension of 29 (122 mmol) in
EtOH (800 mL) and water (150 mL) was added iron powder (673mmol) and NH4Cl (306 mmol).
The mixture was heated under nitrogen at 50oC overnight. Additional iron powder (673 mmol)
and NH4Cl (306 mmol) were added and heating was continued for 45 h. The reaction mixture
was cooled, filtered and concentrated. The residue was dissolved in ethyl acetate and washed
with sodium bicarbonate solution. The organic phase was concentrated to afford the desired
product as a gray solid, which was used in the next step without further purification.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Step 3: 5ꢀChloroꢀ6ꢀiodoꢀ1,3ꢀdihydroꢀ2Hꢀbenzimidazoleꢀ2ꢀthione (31). KOH (238 mmol) in
water (50 mL), followed by carbon disulfide (238 mmol), was added to a solution of 30 (198
mmol) in EtOH (300 mL). The mixture was heated at reflux for 3 h, cooled and filtered. To the
filtrate was added water (300 mL) and then AcOH (25 mL) in water (50 mL). The precipitate
was collected, washed with water and a small amount of EtOH and dried to afford the desired
product as a brown powder, which was used in the next step without further purification.
Step 4: tertꢀButyl 2ꢀ(6ꢀchloroꢀ5ꢀiodoꢀ1Hꢀbenzo[d]imidazolꢀ2ꢀylthio)acetate (32). Cs2CO3 (7.08
mmol), followed by tertꢀbutyl bromoacetate (3.54 mmol), was added to a solution of 31 (3.54
mmol) in THF (20 mL) at 0oC. The reaction was stirred at rt for 0.5 h. Volatiles were removed
and the residue was partitioned between EtOAc and water. Concentration afforded the desired
product as a white power. LCꢀMS: calculated for C13H14ClIN2O2S 423.95, observed m/e 424.8
(M + H)+.
Preparation of intermediate 6-chloro-5-iodo-2-(methylsulfonyl)-1H-benzo[d]imidazole
(33).
ACS Paragon Plus Environment
29